Submerging your mainboard in dihydrogen oxide

[RussianSoldier]

Sure, you could go out and buy hydrofluoroether @ $220/gallon, but has anyone tried using very pure deionized or distilled water?

I'm planning some experiments for this summer, but I'd like to know if anyone has actually had any success...


Edit: Well, my reading on Google isn't turning up very promising results, but I'm interested in any stories...

 

[DerwenArtos12]

haven't heard of someone trying, seems intersting though.

 

[Nebor]

Why not just stomp on it? It would be faster.

 

[Mellman]

there's always the mineral oil solution...

 

[akseli]

Originally posted by: RussianSoldier
Sure, you could go out and buy hydrofluoroether @ $220/gallon, but has anyone tried using very pure deionized or distilled water?

I'm planning some experiments for this summer, but I'd like to know if anyone has actually had any success...


Edit: Well, my reading on Google isn't turning up very promising results, but I'm interested in any stories...

It would work if you could get 100% salt free water (it's the salts that conduct electricity) so like you said it would have to be deionized ... Problem is the water would heap up rapidly and you'd have to have a constant flow ... It would be hard to keep that water clean whilst still being deionized.

 

[DerwenArtos12]

Originally posted by: akseli

Originally posted by: RussianSoldier
Sure, you could go out and buy hydrofluoroether @ $220/gallon, but has anyone tried using very pure deionized or distilled water?

I'm planning some experiments for this summer, but I'd like to know if anyone has actually had any success...


Edit: Well, my reading on Google isn't turning up very promising results, but I'm interested in any stories...

It would work if you could get 100% salt free water (it's the salts that conduct electricity) so like you said it would have to be deionized ... Problem is the water would heap up rapidly and you'd have to have a constant flow ... It would be hard to keep that water clean whilst still being deionized.

IT would have to be compltely sealed airtight. you would have to build a custom enclosure for it that could keep the water moving and keep the entire setup sealed. I am gonna have to just say not worth the effort.

 

[jagec]

Originally posted by: akseli

Originally posted by: RussianSoldier
Sure, you could go out and buy hydrofluoroether @ $220/gallon, but has anyone tried using very pure deionized or distilled water?

I'm planning some experiments for this summer, but I'd like to know if anyone has actually had any success...


Edit: Well, my reading on Google isn't turning up very promising results, but I'm interested in any stories...

It would work if you could get 100% salt free water (it's the salts that conduct electricity) so like you said it would have to be deionized ... Problem is the water would heap up rapidly and you'd have to have a constant flow ... It would be hard to keep that water clean whilst still being deionized.

Considering that 100% pure water conducts, I don't know how well it would work.

 

[Jaxidian]

Originally posted by: jagec
Considering that 100% pure water conducts, I don't know how well it would work.

Pure H2O does not conduct electricity. However, it would be next to impossible for one to a tub of it pure.

 

[mrwxyz]

Originally posted by: Jaxidian

Originally posted by: jagec
Considering that 100% pure water conducts, I don't know how well it would work.

Pure H2O does not conduct electricity. However, it would be next to impossible for one to a tub of it pure.

ya, pure h20 does not conduct. the thing is even if u do manage to get pure h20, particles from mobo would fall into the water and make it conduct. Therefore, u would need pure h20 AND a mobo with a some sort of shielding on it to keep the water from making direct contact. Enter blocks from water cooling.

 

[RS3RS]

For all the time, money, and effort, you wouldn't be getting much out of it in the end. I think you'd be better off spending your time making some sort of direct die phase-change cooling. The end results would be much better, and you'd probably have LESS time and effort in it in the end.

 

[Jaxidian]

Also, from what I've read, even if you were able to successfully do this, some of your stuff will go out before long (as in maybe a month or so). One of two things will happen:

The water will get inside of components, the heat will make the water expand (the water-logged components will expand as well), and something goes. This goes for PCBs, improperly-sealed chips, capacitors, and probably other things.

The adhesive around your CPU die will deteriorate due to water (not sure why but it deteriorates the adhesive stuff that holds your die securely onto the chip package). Once the adhesive is gone, then the 0.13µ leads tend to break. And when leads in a CPU die breaks, that's bad.



Something else that could happen (I've not heard anybody claim this though) is that when the water gets inside of capacitors or inductors, that it throws off the way they work (the dielectric in there is supposed to make them work one way but when water is mixed with that dielectric, they work differently) and then that affects your voltages and stability.

Overall, a bad idea.

 

[Jaxidian]

Originally posted by: RS3RS
For all the time, money, and effort, you wouldn't be getting much out of it in the end. I think you'd be better off spending your time making some sort of direct die phase-change cooling. The end results would be much better, and you'd probably have LESS time and effort in it in the end.

I think, for less money, he should do a water-cooling system, use pure antifreeze (or something that can withstand really cold temps but transfers heat well) and put his radiator and resivoir in a freezer or something. This will give really good results as well (but be sure not to let your waterblock sweat onto your cpu, that would be bad).

 

[pspada]

Um, no. If I were freakin Aquaman, and needed to run my box in my native element, perhaps, otherwise I'll stick with plain ole air, thank you.

 

[DerwenArtos12]

what is the condutivity rating of freon? that is what is used in refridgerators to transfer the heat, and if it doesn't conduct then why not use it instead of water in a regular water cooling set-up?

 

[gsellis]

It was mentioned, but to restate, water is a universal solvant. There may be some things that are soluable that you don't want going into solution...

Also, it would be too expensive to find water that pure. And you could not keep it that pure as stuff will grow in it. You have to have a filter running with the water circulating. Plus, the material it is in may leach.

Not a practical idea, me thinks.

 

[Mark R]

It's unlikely to work well.

Pure water conducts electricity (not very well, but it is definitely a conductor). The other problem is that water is an excellent solvent - and unless anything in contact with the water has been thoroughly washed, then the contamination will dramatically increase conductivity.

Additionally, the small electric currents that will flow through the water (no matter how pure) will cause electrolytic corrosion of exposed metal - these metals will ionise and dissolve, dramatically increasing the conductivity of the water, and making the process worse.

Eventually, the conductivity of the water will rise to the point where it will disrupt circuits and cause malfunction. Even if you had a purifier constantly operating, eventually the solder joints and components will corrode due to electrolysis.

 

[mrwxyz]

Originally posted by: Mark R
It's unlikely to work well.

Pure water conducts electricity (not very well, but it is definitely a conductor). The other problem is that water is an excellent solvent - and unless anything in contact with the water has been thoroughly washed, then the contamination will dramatically increase conductivity.

Additionally, the small electric currents that will flow through the water (no matter how pure) will cause electrolytic corrosion of exposed metal - these metals will ionise and dissolve, dramatically increasing the conductivity of the water, and making the process worse.

Eventually, the conductivity of the water will rise to the point where it will disrupt circuits and cause malfunction. Even if you had a purifier constantly operating, eventually the solder joints and components will corrode due to electrolysis.

i have to disagree....pure water does not conduct....unless i've been lied to all my life

edit: but i do agree about the solvent part completely

 

[Mark R]

Originally posted by: mrwxyz

i have to disagree....pure water does not conduct....unless i've been lied to all my life

edit: but i do agree about the solvent part completely

The conductivity of absolutely pure water is 55 nS/cm. That is an extremely low value, and the currents that flow are likely to be tiny - however, the small size of components on modern electronics as well as the very low signal currents used mean that even these tiny currents could cause disturbance of the circuit.

Certainly, over a period of a few hours, you are likely to get noticable electrolytic corrosion with rapid contamination of the water.

 

[Hyperlite]

well what about this....has anyone attempted to fill a standard water-cooled set up with antifreeze?

 

[gururu]

water conducts because it always contains a small population of hydroxide and hydrogen ion, which are charged. these are components of equilibrium and so cannot be removed by deionization. It is a poor conductor of course as you've all alluded to.
water is the universal solvent, and anything dipped into it is going to shed charged elements into it as pointed out, raising its conductivity with time; enough so that at some timepoint your system is going to short circuit.

 

[gururu]

ooops

also forgot to say that water in contact with exposed circuitry results in the hydrolysis of water and the release of hydrogen, increasing the number of hydroxide ions in solution

 

[DerwenArtos12]

Originally posted by: Hyperlite
well what about this....has anyone attempted to fill a standard water-cooled set up with antifreeze?

Anti-freze does not transfer heat much better than water, what I am still waitin on is if someone has thought of using freon and if it transfers heat better or worse than water and if it is conductive at all. Does anyone know?

 

[RockGuitarDude]

since when is it chem 101 in here? ;-)

 

[EwoutvB]

Originally posted by: Hyperlite
well what about this....has anyone attempted to fill a standard water-cooled set up with antifreeze?

Actualy people have...

The more water VS antifreeze the better cooling, the more antifreeze the crapier the cooling... hehehe

Guys just buy some hydrofluoroether...

 

[NumbersGuy]

How about a tight fitting plexiglass case, wire openings *well* sealed, submerged in H2O + antifreeze circulating through a radiator.

A heat spreader (or heat pipe) over the hot parts would be needed inside & outside the case.